上大学会多大程度上推迟初婚年龄?——基于IV-Tobit模型的估计
|
摘要
在估计高等教育会多大程度上推迟初婚年龄时会同时面临生存数据结构(Duration data)与内生性问题,以往文献无法很好地解决这个问题。基于已有研究方法,本文设计了一种计量策略来解决这一问题。首先将该问题等价转化为研究在给定年龄的条件下高等教育对已婚时间的影响,从而将数据转化为一个单方向截尾的数据结构并使用Tobit模型进行估计。然后用是否在扩招后参加高考作为高等教育的工具变量,用IV-Tobit模型估计高等教育对初婚年龄的推迟效应。估计结果显示,相对于高中毕业生,每多接受1年高等教育平均初婚年龄将推迟1.5年;相应地,大专、本科教育会将初婚年龄推迟4.5和6年。由于女性生育存在一个最佳的年龄阶段,因此该推迟效应可能会严重影响生育率和母婴健康。另外,高等教育对男性初婚推迟的效应大于女性,对非农业户口的影响大于农业户口,在东、中、西部地区没有差异。本文发现这种推迟效应主要是在校接受高等教育带来的顺延,高等教育并没有显著降低接受者的结婚意愿,上大学只是推迟婚姻,而不是终身不婚。人口普查数据显示我国高等教育人口与较低教育人口的最终结婚率几乎没有差异。不同于其他发达国家,我国的高等教育群体仍然具有较高的结婚率。政府应该采取积极措施降低他们步入婚姻的经济成本,延缓这种婚姻推迟的趋势。
It is usually hard to estimate the causal effect of higher education on the age at the first marriage because of the existence of endogeneity problem in a duration data.In this paper,we try to solve this problem using IV-Tobit model to a transformed data.First,we transform the research question on the effect of higher education on the years since the marriage till the survey year,which is censored from the left.Second,we use Tobit and IV-Tobit models to estimate the effect of higher education on years since the marriage.Using the natural experiment,the higher education expansion since 1999,as the instrument,we find that higher education delays the age at the first marriage.Compared to the case had not admitted to colleges/universities,one additional year of higher education postpones the first marriage by 1.5 years.This corresponds 4-6 years postpone of the age of the first marriage due to the higher education.Given that there is an optimal fertility age period for women,this 4-6 years of postpone may have a strong negative effect on fertility and maternal and child health.In addition,we find the delay effect is stronger for the male than female,greater for the non-agriculture Hukou than agricultural Hukou,and has no difference across regions.We also find that this delay effect is mainly due to the extra years of higher education,and does not find higher education changes the willingness to marry.Different from other developed countries,the higher education group of China remains a high marriage rate.The government should take active measures to decrease the economic cost of their marriage and slow down their marriage delay trend.
It is usually hard to estimate the causal effect of higher education on the age at the first marriage because of the existence of endogeneity problem in a duration data.In this paper,we try to solve this problem using IV-Tobit model to a transformed data.First,we transform the research question on the effect of higher education on the years since the marriage till the survey year,which is censored from the left.Second,we use Tobit and IV-Tobit models to estimate the effect of higher education on years since the marriage.Using the natural experiment,the higher education expansion since 1999,as the instrument,we find that higher education delays the age at the first marriage.Compared to the case had not admitted to colleges/universities,one additional year of higher education postpones the first marriage by 1.5 years.This corresponds 4-6 years postpone of the age of the first marriage due to the higher education.Given that there is an optimal fertility age period for women,this 4-6 years of postpone may have a strong negative effect on fertility and maternal and child health.In addition,we find the delay effect is stronger for the male than female,greater for the non-agriculture Hukou than agricultural Hukou,and has no difference across regions.We also find that this delay effect is mainly due to the extra years of higher education,and does not find higher education changes the willingness to marry.Different from other developed countries,the higher education group of China remains a high marriage rate.The government should take active measures to decrease the economic cost of their marriage and slow down their marriage delay trend.
引文
[1]陆杰华,王笑非.20世纪90年代以来我国婚姻状况变化分析[J].北京社会科学,2013,(3):62-72.
[2]张溢.教育是否妨碍了女性的婚姻[D].北京:清华大学,2012.
[3]吴要武,刘倩.高校扩招对婚姻市场的影响:剩女?剩男?[J].经济学(季刊),2015,(1):5-30.
[4]宋健,范文婷.高等教育对青年初婚的影响及性别差异[J].青年研究,2017,(1):1-8.
[5]吴邦国.以“三个代表”重要思想为指导切实做好下岗失业人员再就业工作--在全国再就业工作会议上的报告[J].劳动保障通讯,2003,(2):9-18.
[6]吴要武,赵泉.高校扩招与大学毕业生就业[J].经济研究,2010,(9):93-108.
[7]Yeung W J.China’s Higher Education Expansion and Social Stratification in China[J].Chinese Sociological Review,2013,45(4):54-80.
[8]Gary S Becker.A Theory of Marriage:Part I[J].Journal of Political Economy,1973,81(4):813-846.
[9]Gary S Becker.A Theory of Marriage:Part II[J].Journal of Political Economy,1974,(2):11-26.
[10]Gary S Becker.A Treatise on the Family[M].Harvard University Press,1981:870-871.
[11]Sweeney M M.Two Decades of Family Change:The Shifting Economic Foundations of Marriage[J].American Sociological Review,2002,67(1):132-147.
[12]Oppenheimer V K.A Theory of Marriage Timing[J].American Journal of Sociology,1988,94(3):563-591.
[13]Oppenheimer V K.Women’s Rising Employment and the Future of the Family in Industrial Societies[J].Population&Development Review,1994,20(2):293-342.
[14]Oppenheimer V K.Women’s Employment and the Gain to Marriage:the Specialization and Trading Model[J].Annual Review of Sociology,1997,23(1):431-453.
[15]Goldscheider F K,Waite L J.Sex Differences in the Entry into Marriage[J].American Journal of Sociology,1986,92(1):91-109.
[16]刘爽,高华.我国人口初婚风险影响因素及其变迁--基于CGSS2010的研究[J].南方人口,2015,(1):1-14.
[17]杜本峰.时间史分析及其应用[M].北京:经济科学出版社,2008:23-31.
[18]Cox D.Regression Models and Life Tables[J].Journal of the Royal Statistical Society Series B,1972,34(2):187-220.
[19]梁同贵.婚前同居与初婚离婚风险--基于CFPS2010数据的分析[J].南方人口,2017,32(4):45-55.
[20]历年全国高考报名人数与录取率[Z/OL].(2011-05-04).http://edu.people.com.cn/GB/116076/14545024.html.
[21]韦艳,董硕,姜全保.中国初婚模式变迁--基于婚姻表的分析[J].人口与经济,2013,(2):21-28.
[22]Tobin J.Estimation of Relationships for Limited Dependent Variables[J].Econometrica,1958,26(1):24-36.
[23]周华林,李雪松.Tobit模型估计方法与应用[J].经济学动态,2012,(5):105-119.
[24]Newey W K.Efficient Estimation of Limited Dependent Variable Models with Endogenous Explanatory Variables[J].Journal of Econometrics,1987,36(3):231-250.
(1)CGSS没有公布2014年数据。
(1)最早提出用性别角色分工理论解释婚姻形成的是社会学者Parsons(1949),Becker的理论是在其理论上进行了加工和阐释。
(1)2010年人口普查数据中结婚比例最低的是文盲。2010年35岁、40岁的文盲结婚比例分别为80%、85%,这个群体结婚比例偏低主要是因为人口性别比例失衡造成的,并非不愿结婚。
(1)基础风险函数一般都是未知的,因此Cox模型适用于来进行影响因素分析,不适合估计生存率。
(1)后面我们也加入了2008年数据进行回归,估计结果是一致的。2006年及以前的数据由于最早受影响的本科生毕业才3年左右,不足以进行婚姻行为分析。
(1)由于没有结婚前户口或者出生时父母的户口信息,因此这里的户口指的是当前本人户口信息。
(1)近十年《中国教育统计年鉴》的数据显示,毕业生中普通专科和普通本科的比例在1∶1左右。因此,按照1∶1的比例计算出来上大学相对于高中毕业生平均初婚推迟为(4.5+6.0)/2,约为5.25年左右。当然,相比于高中及以下的群体,初婚推迟要大于5.25年。
[2]张溢.教育是否妨碍了女性的婚姻[D].北京:清华大学,2012.
[3]吴要武,刘倩.高校扩招对婚姻市场的影响:剩女?剩男?[J].经济学(季刊),2015,(1):5-30.
[4]宋健,范文婷.高等教育对青年初婚的影响及性别差异[J].青年研究,2017,(1):1-8.
[5]吴邦国.以“三个代表”重要思想为指导切实做好下岗失业人员再就业工作--在全国再就业工作会议上的报告[J].劳动保障通讯,2003,(2):9-18.
[6]吴要武,赵泉.高校扩招与大学毕业生就业[J].经济研究,2010,(9):93-108.
[7]Yeung W J.China’s Higher Education Expansion and Social Stratification in China[J].Chinese Sociological Review,2013,45(4):54-80.
[8]Gary S Becker.A Theory of Marriage:Part I[J].Journal of Political Economy,1973,81(4):813-846.
[9]Gary S Becker.A Theory of Marriage:Part II[J].Journal of Political Economy,1974,(2):11-26.
[10]Gary S Becker.A Treatise on the Family[M].Harvard University Press,1981:870-871.
[11]Sweeney M M.Two Decades of Family Change:The Shifting Economic Foundations of Marriage[J].American Sociological Review,2002,67(1):132-147.
[12]Oppenheimer V K.A Theory of Marriage Timing[J].American Journal of Sociology,1988,94(3):563-591.
[13]Oppenheimer V K.Women’s Rising Employment and the Future of the Family in Industrial Societies[J].Population&Development Review,1994,20(2):293-342.
[14]Oppenheimer V K.Women’s Employment and the Gain to Marriage:the Specialization and Trading Model[J].Annual Review of Sociology,1997,23(1):431-453.
[15]Goldscheider F K,Waite L J.Sex Differences in the Entry into Marriage[J].American Journal of Sociology,1986,92(1):91-109.
[16]刘爽,高华.我国人口初婚风险影响因素及其变迁--基于CGSS2010的研究[J].南方人口,2015,(1):1-14.
[17]杜本峰.时间史分析及其应用[M].北京:经济科学出版社,2008:23-31.
[18]Cox D.Regression Models and Life Tables[J].Journal of the Royal Statistical Society Series B,1972,34(2):187-220.
[19]梁同贵.婚前同居与初婚离婚风险--基于CFPS2010数据的分析[J].南方人口,2017,32(4):45-55.
[20]历年全国高考报名人数与录取率[Z/OL].(2011-05-04).http://edu.people.com.cn/GB/116076/14545024.html.
[21]韦艳,董硕,姜全保.中国初婚模式变迁--基于婚姻表的分析[J].人口与经济,2013,(2):21-28.
[22]Tobin J.Estimation of Relationships for Limited Dependent Variables[J].Econometrica,1958,26(1):24-36.
[23]周华林,李雪松.Tobit模型估计方法与应用[J].经济学动态,2012,(5):105-119.
[24]Newey W K.Efficient Estimation of Limited Dependent Variable Models with Endogenous Explanatory Variables[J].Journal of Econometrics,1987,36(3):231-250.
(1)CGSS没有公布2014年数据。
(1)最早提出用性别角色分工理论解释婚姻形成的是社会学者Parsons(1949),Becker的理论是在其理论上进行了加工和阐释。
(1)2010年人口普查数据中结婚比例最低的是文盲。2010年35岁、40岁的文盲结婚比例分别为80%、85%,这个群体结婚比例偏低主要是因为人口性别比例失衡造成的,并非不愿结婚。
(1)基础风险函数一般都是未知的,因此Cox模型适用于来进行影响因素分析,不适合估计生存率。
(1)后面我们也加入了2008年数据进行回归,估计结果是一致的。2006年及以前的数据由于最早受影响的本科生毕业才3年左右,不足以进行婚姻行为分析。
(1)由于没有结婚前户口或者出生时父母的户口信息,因此这里的户口指的是当前本人户口信息。
(1)近十年《中国教育统计年鉴》的数据显示,毕业生中普通专科和普通本科的比例在1∶1左右。因此,按照1∶1的比例计算出来上大学相对于高中毕业生平均初婚推迟为(4.5+6.0)/2,约为5.25年左右。当然,相比于高中及以下的群体,初婚推迟要大于5.25年。
© 2004-2018 中国地质图书馆版权所有 京ICP备05064691号 京公网安备11010802017129号 地址：北京市海淀区学院路29号 邮编：100083 电话：办公室：(+86 10)66554848；文献借阅、咨询服务、科技查新：66554700 |